Modified Vaccinia Ankara (MVA) virus is related to vaccinia virus, a member of the genera Orthopoxvirus in the family of Poxyiridae. MVA was generated by 516 serial passages on chicken embryo fibroblasts of the Ankara strain of vaccinia virus (CVA) (for review see Mayr, A., et al. Infection 3, 6-14 [1975]). As a consequence of these long-term passages, the resulting MVA virus deleted about 31 kilobases of its genomic sequence and, therefore, was described as highly host cell restricted to avian cells (Meyer, H. et al., J. Gen. Virol. 72, 1031-1038 [1991]). It was shown in a variety of animal models that the resulting MVA was significantly avirulent (Mayr, A. & Danner, K. [1978] Dev. Biol. Stand. 41: 225-34). Additionally, this MVA strain has been tested in clinical trials as a vaccine to immunize against the human smallpox disease (Mayr et al., Zbl. Bakt. Hyg. I, Abt. Org. B 167, 375-390 [1987], Stickl et al., Dtsch. med. Wschr. 99, 2386-2392 [1974]). These studies involved over 120,000 humans, including high-risk patients, and proved that compared to vaccinia based vaccines, MVA had diminished virulence or infectiousness while it induced a good specific immune response.
In the following decades, MVA was engineered for use as a viral vector for recombinant gene expression or as a recombinant vaccine (Sutter, G. et al. [1994], Vaccine 12: 1032-40).
In this respect, it is most astonishing that even though Mayr et al. demonstrated during the 1970s that MVA is highly attenuated and avirulent in humans and mammals, some recently reported observations (Blanchard et al., 1998, J Gen Virol 79, 1159-1167; Carroll & Moss, 1997, Virology 238, 198-211; Altenberger, U.S. Pat. No. 5,185,146; Ambrosini et al., 1999, J Neurosci Res 55(5), 569) have shown that MVA is not fully attenuated in mammalian and human cell lines since residual replication might occur in these cells. It is assumed that the results reported in these publications have been obtained with various known strains of MVA since the viruses used essentially differ in their properties, particularly in their growth behavior in various cell lines.
Growth behavior is recognized as an indicator for virus attenuation. Generally, a virus strain is regarded as attenuated if it has lost its capacity or only has reduced capacity to reproductively replicate in host cells. The above-mentioned observation, that MVA is not completely replication incompetent in human and mammalian cells, brings into question the absolute safety of known MVA as a human vaccine or a vector for recombinant vaccines.
Particularly for a vaccine, as well as for a recombinant vaccine, the balance between the efficacy and the safety of the vaccine vector virus is extremely important.